Web Newsletter
Front Page Livingston Caltech


"If a tree falls in a forest..." Progress on Seismic Pre-isolation

"If a tree falls in a forest..." Progress on Seismic Pre-isolation

- Contributed by David Shoemaker

The two LIGO Observatory sites at Hanford, Washington, and Livingston, Louisiana, were chosen for their suitability for the endeavor, and prime among the criteria was the need for a quiet seismic environment. Most times, both regions are really very quiet. But at other times, not so. (See this article from a past LIGO Newsletter on the subject of the present seismic spectrum; and here is a review of measurements.) The bottom line is that the combination of the seismic noise at the sites (especially from 1-3 Hz), the rather high Q of our seismic isolation system at the solid-body resonant frequencies, and the finite dynamic range of the suspension controllers means that we have difficulty locking the interferometers when there is a lot of human activity.

This high level of seismic noise, and LIGO's sensitivity to it, can be a problem mainly at the Livingston Observatory. This article's title, "If a tree falls..." is a reference to the fact that a lot of the seismic activity at Livingston stems from the commercial logging done in the area and which has recently increased, particularly near one of the end test masses. As a result, the duty cycle on that instrument is not as high as we would want. It's also likely, as our other noise sources are suppressed, that we will eventually find that the large seismic noise at frequencies up to 10 Hz or so will be "upconverted" through non-linear effects into our gravitational-wave band. This is another incentive for suppressing this noise.

Our solution to this problem takes advantage of the Advanced LIGO Seismic Isolation design, currently in prototyping. (See the Seismic Isolation for Advanced LIGO document for more details on this design.) It uses a combination of in-vacuum "active" servo-controlled quieted stages, as well as an external seismic pre-isolator. It is this last element that looks to be a very effective way to quiet down the seismic noise troubling us. It has a variety of virtues:

Once we identified the difficulty with the seismic noise, we put development of the pre-isolator on the proverbial "front burner," with the objective of getting it installed as quickly as possible. The development of the concept, and a realization using a hydraulic actuator, has come from our collaborators at Stanford (Take a look at this Work on Hydraulics document for greater detail). We are also pursuing a design with electromagnetic actuators at MIT.

We are now at the point of putting together some of the prototypes for a test we will perform of the LASTI installation at MIT. Some photos below show where we are.

Figure 1.

To briefly describe the design: the LIGO Seismic Isolation is supported upon four "blue piers" (even though they have come out as gray in the left-side illustration above). In the initial LIGO design, scissors jacks and translation tables allow some adjustment of the quasi-static position of the seismic isolation (and optics attached to it). On the end test masses, there is also a fine actuator designed for slow corrections (tidal, microseismic) along the optical path. These are currently being extended to use at higher frequencies, as a quick (and remarkably successful!) means to reduce the excess noise.

The pre-isolator design calls for removal of the scissors jack, translation table, and some other hardware between the top of the blue piers and the actual structure of the seismic isolation (shown in the middle illustration above), and its replacement with the new pre-isolator, one at each of the four corners. This pre-isolator has to support the very significant weight of the seismic isolation system, so some very substantial springs are needed to off-load this weight. The two photos below show prototype springs. In the left-side image, you can see that the spring has clockwise and counterclockwise segments to eliminate torques upon extension. A very nice "clean" alternative design, which we'll use from here onward, is shown in the image at right. This machined spring is easier to make reliably (and is very pretty, to boot!).

Figure 2. Figure 3.

The mechanical design of the structure, from Ken Mason at MIT along with the rest of the Seismic team, uses two springs at a 60 degree angle to achieve its purpose, as shown in the next photo below. Ken admits this concept was inspired by a Harley V-twin engine. There are actuators (hydraulic or electromagnetic) and seismometers and position detectors, which effectively allow the piers to move with seismic noise while holding the suspension point of the seismic isolation system stationary in inertial space.

Figure 4.

This next photo below shows one critical step in the construction of one of the assemblies. Note that, in keeping with the engine inspiration, an actual engine hoist is the crane of preference! From left to right, we see Ken Mason, David Shoemaker, Myron MacInnis, Megan Goldman, Dennis Coyne, and Gregg Harry (how many GW experimentalists does it take to...?) all helping to lower the block which is supported by the springs into the outer housing. Two of those yellow springs will be inserted into the two cylinder bores and attached to the block, and then the unit is moved into place on a pier.

Figure 5.

And finally, we see one corner of the HAM installed electromagnetic pre-isolator (MEPI). The others are being installed as I type!

Figure 6.

Once we have a complete set of four together and installed, we will start the process of characterizing the resonant frequencies of the system, understanding the compromises from the imperfect alignments of actuators and sensors, and begin to get the control system functioning. We will first test the variant using electromagnetic actuators on a HAM chamber. Then, a bit later this fall, we will test the baseline version using the hydraulic actuator on a BSC chamber. And even though we are not exactly thrilled to be in need of this pre-isolator, or to have to spend a lot of time on this seismic noise problem while other R&D work is put on temporary hold, we are in fact delighted to see the LASTI installation prove its value: this is currently the only place in the world where we could be testing these isolators without interfering with the commissioning and experiments going on at the observatories. So stay tuned--we should have some initial results in a few months!

    Onwards! to the LIGO MIT Home Page     Or  Return! to the Newsletter Front Page